Performance analysis of normalized least mean p-norm lattice algorithm for alpha-stable processes

The existence of impulsive noise with α-stable distributions has been addressed in many applications. We present two new adaptive algorithms using the lattice structure for α-stable processes. In these algorithms fractional lower order moments of residual errors are used to update the filter coefficients. Based on the empirical results, the proposed algorithms show superior convergence speed over presently available techniques in parameter estimation of α-stable processes. The problem of misalignment of the mean of the estimated parameters with respect to the true values is also addressed